PEDIATRICS Vol. 106 No. 2
August 2000,
p. e25
ELECTRONIC ARTICLE:
Early Cognitive and Motor Development Among Infants Born to Women
Infected With Human Immunodeficiency Virus
Cynthia Chase, PhD*,
Janice Ware, PhD
,
Joan Hittelman, PhD§,
Ileana Blasini, MD,
MPH
;
Renee Smith, PhD¶,
Antolin Llorente, PhD#,
Elizabeth Anisfeld, PhD**,
Clemente Diaz, MD
,
Mary Glenn Fowler, MD, MPH§§,
Jack Moye, MD||,
Leslie I. Kaligh, DSc||, and
for the
Women and Infants Transmission Study Group
From the * Department of Pediatrics, Division of Infectious
Disease, Boston Medical Center, Boston University School of Medicine,
Boston, Massachusetts; Department of Pediatrics, Division of
Infectious Diseases, Children's Hospital, Harvard Medical School,
Boston, Massachusetts; § Health Science Center, Brooklyn and State
University of New York, Brooklyn, New York; Department of
Pediatrics, University of Puerto Rico, San Juan, Puerto, Rico;
¶ Department of Pediatrics, University of Illinois, Urbana,
Illinois; # Department of Pediatrics, Baylor College of Medicine,
Houston, Texas; ** Department of Pediatrics, Columbia-Presbyterian
Hospital, New York, New York; Division of Acquired
Immunodeficiency Syndrome, National Institute of Allergy and Infectious
Diseases, Bethesda, Maryland; §§ Pediatric Adolescent and
Maternal AIDS Branch, National Institute of Child Health and Human
Development, Washington, DC; and || New England Research
Institutes, Watertown, Massachusetts.
 |
ABSTRACT |
Objective. To examine the frequency,
timing, and factors associated with abnormal cognitive and motor
development during the first 30 months of life in infants born to women
infected with human immunodeficiency virus type 1 (HIV-1).
Methods. Serial neurodevelopmental assessment was
performed with 595 infants born to women infected with HIV-1 in a
multicenter, prospective, natural history cohort study. Survival
analysis methods were used to evaluate 6 outcome events related to
abnormal cognitive and motor growth (time to confirmed drop of 1 SD,
time to first score <69, and time to confirmed drop of 2 SD) in Bayley
Scales of Infant Development Mental Developmental Index (MDI) and
Psychomotor Developmental Index (PDI) scores among infected
(n = 114) and uninfected (n = 481) infants. Proportional hazards modeling was used to evaluate the
effects of HIV infection status, prematurity, prenatal exposure to
illicit drugs, maternal educational attainment, and primary language.
Results. HIV-1 infection was significantly associated with
increased risk for all outcome events related to abnormal mental and
motor growth. Differences between infected and uninfected infants were apparent by 4 months of age. Prematurity was associated with increased risk for MDI <69 and PDI <69. Maternal education of <9 completed years was associated with increased risk for MDI <69. Neither prenatal
exposure to illicit drugs nor primary language other than English was
associated with abnormal development.
Conclusion. A significant proportion of infants with HIV-1
infection show early and marked cognitive and motor delays or declines
that may be important early indicators of HIV disease progression. These abnormalities are independent of other risk factors for developmental delay.
Key words:
human immunodeficiency virus,
child development.
Neurological and developmental signs are often early
markers of human immunodeficiency virus type 1 (HIV-1) disease in
infants that may precede other signs of disease
progression.1 Numerous observational data describe the
frequency of severe neurodevelopmental abnormalities in pediatric HIV
disease. A prospective study of French children with congenital or
perinatal HIV-1 infection reported a 19% rate of significant central
nervous system impairment.2 The European Collaborative
Study reported a 13% rate of serious neurological signs and symptoms
in infected children and a much higher rate (31%) in children with
acquired immunodeficiency syndrome (AIDS) or AIDS-related
complex.3 A recent multicenter epidemiologic study in the
United States showed HIV-1 encephalopathy diagnosed in 23% of children
with perinatally acquired AIDS.4 HIV encephalopathy was
present in 15% of US pediatric AIDS cases reported in
19945 and 18% in 1995.6 However, the
frequency and timing of early and less severe neurodevelopmental effects of HIV infection, and the interaction with other known risk
factors for developmental impairment, are not well-characterized.
Previous published analyses of neurodevelopmental abnormality in
children with HIV-1 disease have been cross-sectional in design or had
small sample size, reporting differences between HIV-infected and
HIV-exposed but HIV-uninfected infants in mean scores on
neurodevelopmental measures at specific ages.7-9 These
studies generally have not adjusted for other important risk factors
related to neurodevelopmental abnormality besides HIV status. The Women
and Infants Transmission Study (WITS) is an ongoing observational study
of maternal-infant HIV transmission and outcome in women infected with
HIV-1 and their children. Accrual began in 1989, and the first infant
to be enrolled in WITS was born in 1990. The prospective,
systematically collected neurodevelopmental data now available from
this study provide an opportunity to examine longitudinal patterns of
growth in a multivariate context. The study objectives were to examine
the frequency and timing of abnormality in cognitive and motor
development in infants born to HIV-infected women and to identify risk
factors other than HIV infection that are associated with markers of
neurodevelopmental abnormality in these children during the first 30 months of life.
| |
METHODS |
Population Studied
Eligible subjects were all singleton infants born to women with
HIV-1 infection 15 to 44 years old who entered WITS, a prospective natural history cohort study of maternal-infant HIV infection conducted at 6 clinical centers in the mainland United States and
Puerto Rico. Maternal HIV infection was documented by repeated positive
results in the US Food and Drug Administration-licensed serologic
enzyme immunoassay for HIV-1 antibody with independent supplemental
confirmatory test or diagnosis of AIDS as defined by the Centers for
Disease Control and Prevention. Analyses were restricted to data
collected through the 30-month visit for the 595 such infants in whom
the presence or absence of HIV-1 infection had been established by
study protocol and who completed at least 1 neurodevelopmental
assessment. HIV-1 infection was defined by the presence of 2 or more
cultures of peripheral blood mononuclear cells positive for HIV-1
(n = 114). Absence of HIV infection (n = 481) was defined by 2 or more cultures negative for HIV-1 at 1 month
of age or older, with 1 or more cultures negative for HIV at 6 months
of age or older, and no positive culture results. Products of multiple
births and infants born to mothers already enrolled were excluded. Only
subjects with 3 or more neurodevelopmental assessments were included in
analyses of confirmed drops in mental or motor scores (infected,
n = 77 and uninfected, n = 344).
Thirteen subjects died before 30 months old. All 13 of these infants
were HIV-infected. Of these 13 infants, 6 had Metal Developmental Index (MDI) scores <3, and 7 had Psychomotor Developmental Index (PDI) scores <3.
Procedures
Informed consent was obtained from each mother during pregnancy
for herself and the fetus and again after the child's birth for
enrollment of the child. Neurodevelopmental assessment was performed at
4, 9, 12, 15, 18, 24, and 30 months of age. The Bayley Scales of Infant
Development (BSID),10 was administered in strict
accordance with published test instructions and standard test
procedures by trained examiners uninformed of subject protocol-defined
infection status.
The BSID was selected as a widely used and accepted, well-standardized,
dependable multipart instrument for evaluation of developmental status
in the first 21/2 years of life. The instrument has 163 items on
the Mental Scale and 81 items on the Motor Scale. The Mental Scale is
designed to assess sensory-perceptual acuities, discriminations, and
the ability to respond to these; early acquisition of object constancy
and memory, learning, and problem-solving ability; vocalizations and
beginnings of verbal communications; and early evidence of ability to
form generalizations and classifications. Results are expressed as a
standard score, the MDI. The Motor Scale is designed to provide a
measure of the degree of control of the body, coordination of the large
muscles, and finer manipulative skills of the hands and fingers. It is
specifically directed toward behaviors reflecting motor coordination
and skills. Results are expressed as a standard score, the PDI. The
standard scores are normalized for each of 14 age groups to have a mean
value of 100 and a standard deviation (SD) of 16. The standard score
range of 50 to 150 covers >3 SD on either side of the average MDI or PDI for each age.
Intertester reliability was maintained by centralized training for WITS
examiners and supervisors, periodic review and clarification of scoring
criteria, and selected review of videotapes of the assessments.
Neurodevelopmental assessments were uniformly scheduled to precede any
invasive or painful procedures, such as phlebotomy, and postponed in
instances of acute illness. Assessments were conducted in the primary
language used in the subject's home, by an examiner fluent in the
primary language or through an interpreter who assisted with the
assessment. The raw and standard scores of the BSID Mental and
Psychomotor Scales were recorded at each assessment for each subject.
Any examination limiting factors, such as poor cooperation or poor
intelligibility on the child's part, were noted and recorded.
For MDI and PDI scores <50, an extrapolated score was obtained using
tables by Naglieri11 of statistically derived expected
scores for the BSID. The minimum score using these extrapolated scores
is 28 rather than 50. For premature infants (<37 completed weeks of
gestation), actual chronologic ages without adjustment were used in
calculating MDI and PDI. Prematurity was incorporated as a covariate in
the modeling process as a dichotomous variable. Gestational age was
determined by a combination of prenatal ultrasonography, physical
examination of uterine fundal height, and menstrual history. Prenatal
drug exposure was defined by maternal prenatal use of illicit drugs (opiates, cocaine, or other injectables) or methadone, as assessed by a
combination of interview and urine toxicologic studies. The measure of
maternal disease progression was the average of all maternal CD4
measures available during pregnancy. Other covariates and
sociodemographic characteristics, including primary language and
maternal education, were assessed by trained interviewers using uniform
questionnaires and interview techniques.12 Informed
consent was obtained for all women and infant subject participants
according to local institutional review board and federal
guidelines13 and regulations.14 Consent for
infant enrollment was obtained from the legal guardian within 7 days
after birth.
Statistical Methods
Mean MDI and PDI scores were calculated for each visit by HIV
status. Survival analysis and proportional hazards modeling techniques
were used to evaluate 3 different types of abnormal cognitive (measured
by MDI) or motor (measured by PDI) developmental outcome events in
order of increasing severity: a 16-point (1 SD) confirmed drop in
scaled score, first occurrence of a scaled score <69 (2 SD or more
below mean), and a 32-point (2 SD) confirmed drop in scaled score. The
BSID was constructed psychometrically with a mean of 100 and SD of 16, thus defining a score of 
69 as below the normal range. To ensure that
our results were not driven by arbitrary choices in the endpoint
definitions, we considered both drops of 1 and 2 SDs, with the
requirement of confirmation at a subsequent visit. In recently
published and ongoing trials by the Pediatric AIDS Clinical Trials
Group, a drop in MDI or PDI of 2 SD of 1 occasion or 1 SD
confirmed on a second test are used as neuropsychologic endpoints to
define treatment failure.15-17 Confirmed drops
were defined as declines of at least the specified magnitude (16 or 32 points) below any previous score for a given subject, with the next
consecutive score for that subject also being at least the specified
magnitude (16 or 32 points) below any score previous to the 1 representing the initial drop. Subjects whose maximum MDI or PDI scores
precluded the possibility of dropping a full 16 or 32 points were
excluded from analyses of drops in those specific scores but were
included in the analyses of scores <69. There were 6 such exclusions
from 3 infected children.
Kaplan-Meier estimates of the cumulative probability of each 1 of the 3 outcome events across time for each of the 2 developmental indexes (MDI
and PDI) were generated for each of the covariates (HIV infection,
prematurity, prenatal drug exposure, maternal educational attainment
<9 completed years, and primary language other than English) and its
complement and were compared using the log-rank test. The
simultaneous effects of all 5 covariates on the risk of reaching an
outcome event were assessed using a multivariable Cox proportional
hazards regression model. Additional analyses were performed excluding
all children whose gestational age was <34 weeks and adding maternal
CD4% as a covariate in the regression model.
| |
RESULTS |
Description of Sample
Sociodemographic and maternal characteristics of infant subjects
were similar for the overall sample (n = 595) and for
the subgroup with 3 or more assessments (n = 421) and
are summarized in Table 1 by infection
status. HIV-infected subjects more frequently experienced premature
birth, prenatal drug exposure, and lower maternal educational
attainment. A higher proportion of biological mothers of HIV-infected
subjects were white and a higher proportion of families of HIV-infected
subjects spoke English as the primary language, compared with those of
HIV-uninfected subjects. Not unexpectedly, immune function, as measured
by the maternal CD4%, was lower for mothers of HIV-infected subjects, compared with mothers of HIV-uninfected subjects.
View this table:
[in this window]
[in a new window]
|
TABLE 1
Description of Sample of Children With More Than or Equal to One and
More Than or Equal to Three Examinations by HIV Status
|
|
By the closing date for data analysis, 534 of the study children (90%)
had reached their first birthday. Of these, an MDI and/or PDI score was
obtained for 390 (73%). A total of 410 had reached their second
birthday. Of these, 277 (68%) provided a developmental score.
Table 2 displays the mean MDI and PDI
scores by visit for HIV-infected and HIV-negative children. Unadjusted
mean scores were significantly lower in the HIV-negative group at every visit up to and including the 24-month visit.
View this table:
[in this window]
[in a new window]
|
TABLE 2
Mean MDI and PDI Scores at Each Visit by HIV Status, With P
Value for Statistical Significance of the Difference in Mean Scores
|
|
Multivariate Analysis of Neurodevelopmental Indices
HIV-1 infection was a significant predictor for all 6 neurodevelopmental outcome events when considered alone or adjusted for
the other covariates. Of the other 4 covariates, prematurity and
maternal education showed significant association with the occurrence
of any of the outcome events related to mental or motor growth. Neither
prenatal exposure to hard drugs nor primary language in the home showed
significant associations with the declines or deficits in the mental or
motor scores examined. Interactions among covariates were not
significant.
Confirmed Drops of 1 SD in Mental/Motor Scores
Results of the proportional hazards analysis considering the
simultaneous effects of all covariates on risk for 16-point (1 SD)
confirmed drops in MDI or PDI are shown in Table
3. HIV-1 infection was associated with
elevated risk both for mental (MDI, risk ratio [RR]: 1.50; 95% CI
[confidence interval]: 1.03-2.18; P = .034) or motor
(PDI, RR: 1.66; 95% CI: 1.12-2.45; P = .011) outcome
events of this type. Thus, given 2 subjects of the same age who have
not yet experienced a 16-point (1 SD) confirmed drop in cognitive or
motor performance as measured by the BSID, the HIV-infected subject is
one half to two thirds again more likely to experience such a drop in
the next time interval, compared with the HIV-uninfected subject, all
other things being equal. Unexpectedly, premature infants were less
likely to experience abnormal motor (PDI) outcome events of this type
(RR: .51; 95% CI: .31-.85; P = .009), compared with
term infants. No increased risk for reaching a 16-point confirmed drop
in MDI or PDI was associated with prenatal drug exposure, low maternal
educational attainment, or non-English primary language.
View this table:
[in this window]
[in a new window]
|
TABLE 3
Multivariable Cox Proportional Hazards Models Conditional RR and 95%
Confidence Limits for One SD Confirmed Drops in MDI and PDI Scores
|
|
Figures 1 and
2 depict the cumulative probability of
experiencing a 16-point confirmed drop in MDI or PDI from 4 to 30 months of age in HIV-infected, compared with HIV-uninfected subjects.
HIV-infected subjects were more likely than uninfected subjects to
experience either of these 2 outcome events indicating deceleration of
developmental growth after 9 months of age. However, these outcomes
were not specific to the HIV-infected group. The 2-year incidence of
a16-point confirmed drop in MDI was 53.8% in HIV-infected subjects,
compared with 39.5% among HIV-uninfected subjects (log-rank
P = .01) by 24 months of age. Similarly, for PDI, the
2-year incidence of a 16-point confirmed drop was 49.5% among
HIV-infected subjects, compared with 36.8% among HIV-uninfected subjects (log-rank P = .06). Consistent with
multivariate analysis, full-term infants exhibited a 41.7% probability
by 24 months old of a confirmed drop in PDI, compared with 27.9% among
the premature subjects (log-rank P = .01).
Mental/Motor Scores Less Than 69
Table 4 shows the results of
proportional hazards analysis for occurrence of mental (MDI) and motor
(PDI) scores 2 or more SDs below the mean (MDI or PDI <69). Both HIV-1
infection and prematurity were strongly associated with elevated risk
of reaching these abnormal cognitive or motor outcomes. RRs for MDI <69 were 4.88 (95% CI: 2.94-8.12; P = .0001) for
HIV-infected compared with HIV-uninfected subjects, and 4.03 (95% CI:
2.44-6.66; P = .0001) for premature compared with term
infants. Similarly, RRs for PDI <69 were 4.75 (95% CI: 2.79-8.06;
P = .0001) and 3.37 (95% CI: 1.98-5.73;
P = .0001) for HIV-infected subjects and for infants
born prematurely, respectively. Higher maternal educational attainment
showed an association with reduced risk of developing MDI <69 (RR:
.52; 95% CI: .28-.99; P = .032). No increase in risk of reaching abnormal cognitive or motor outcome events of this type was
associated with prenatal drug exposure or with non-English primary
language.
View this table:
[in this window]
[in a new window]
|
TABLE 4
Multivariable Cox Proportional Hazards Models Conditional RR and 95%
Confidence Limits for MDI and PDI Scores Less Than 69
|
|
HIV-infected subjects were more likely than uninfected subjects to
exhibit MDI or PDI scores <69 (2 SD below the mean) at all ages
examined. Figures 3 and
4 depict the cumulative probability of
occurrence of this type of abnormal cognitive (MDI) or motor (PDI)
outcome event from 4 to 30 months of age in HIV-infected subjects,
compared with HIV-uninfected subjects. Early, sustained, and increasing
differences between the HIV-infected and HIV-uninfected subjects were
observed at the earliest time point examined and all subsequent time
points. The incidence of MDI <69 was 16.5% by 12 months and 35.6% by
24 months for HIV-infected infants, compared with 3.2% by 12 months and 8.2% by 24 months among HIV-uninfected infants (log-rank
P = .001).
Confirmed Drops of 2 SD in Mental or Motor Scores
Results of the proportional hazards analysis considering the
simultaneous effects of all covariates on risk for 32-point (2 SD)
confirmed drops in MDI or PDI are shown in Table
5. HIV-1 infection was strongly
associated with elevated risk for abnormal cognitive (MDI, RR: 2.40;
95% CI: 1.27-4.56; P = .007) and motor (PDI, RR:
3.81; 95% CI: 1.93-7.54; P = .0001) outcome events of
this type. Thus, given 2 subjects of the same age who have not yet
experienced a 32-point (2 SD) confirmed drop in cognitive or motor
performance as measured by the BSID, a subject with HIV-1 infection is
more than twice as likely to experience it in the next time interval
than an uninfected subject, all other things being equal. None of the
other covariates examined (gestational age, prenatal drug exposure,
maternal educational attainment, or primary language) were
significantly related to risk for reaching a 32-point confirmed drop in
MDI or PDI.
View this table:
[in this window]
[in a new window]
|
TABLE 5
Multivariable Cox Proportional Hazards Models Conditional RR and 95%
Confidence Limits for Two SD Confirmed Drops in MDI and
PDI Scores
|
|
Figures 5 and
6 depict the cumulative probability of
experiencing a 32-point confirmed drop in MDI or PDI from 4 to 30 months of age in HIV-infected subjects, compared with HIV-uninfected
subjects. Differences were apparent beginning at 9 months old.
HIV-infected subjects were more likely than HIV-uninfected subjects to
experience this type of outcome event at all subsequent ages examined.
The proportions of HIV-infected subjects compared with HIV-uninfected
subjects experiencing a 32-point confirmed drop in cognitive (MDI) were
6.6% versus 1.5% by 12 months and 20.7% versus 10.4% by 24 months
(log-rank P = .01).
Table 6 shows the gestational age
distribution by HIV status. The number of infants with gestational age
<34 weeks was 19 (23%) and 9 (31%), for HIV-infected and
HIV-uninfected infants, respectively. We repeated the multivariate
analyses after excluding the infants with gestational age <34 weeks.
The associations between HIV status and developmental outcomes changed
only slightly when the premature infants were excluded. Excluding
infants with gestational age <34 weeks did not substantively alter the
results.
When prenatal maternal CD4% was added to the multivariate model, the
results changed slightly. Table 7 shows
the result of the proportion hazards models containing all covariates and maternal CD4% for all 6 outcome events. After adjusting for maternal CD4%, HIV status was still significantly associated with all
PDI outcomes and MDI scores of <69. HIV status was associated with an
MDI 1 SD drop (P = .09) and an MDI 2 SD drop
(P = .11), but the association was no longer
statistically significant.
| |
DISCUSSION |
Analysis of prospective neurodevelopmental data from the WITS
cohort of infants born to women infected with HIV-1 suggests that, with
or without accounting for other factors known or suspected to have
adverse influence on developmental growth, several types of abnormal
cognitive and motor neurodevelopmental outcome events occur earlier and
at significantly increased frequency throughout the first 21/2
years of life among HIV-infected infants, compared with HIV-exposed but
HIV-uninfected infants. These findings are in essential agreement with
previous studies.2-4,7,8 Using a cross-sectional or
externally defined measure of abnormality (standard scores 2 SD or more
below age-specific normative mean) significant cognitive and motor
deficits already were evident among an increased proportion of
HIV-infected infants, compared with HIV-uninfected infants, by 4 months
of age
the earliest time point examined in this study. Using
individual growth curve measures of abnormality (16-point/1 SD or
32-point/2 SD confirmed drops within a given subject's scores compared
with previous scores), HIV infection likewise was strongly associated
with increased frequency of abnormal cognitive or motor
neurodevelopment outcome events beginning at 9 months of agethe
earliest time point at which such an event could occur in the study.
Overall, the data indicate that significant proportions of HIV-infected
infants experience delays or declines in cognitive and/or motor
development, which may be important early indicators of HIV disease
progression.
Our investigation of the impact of HIV disease on early
neurodevelopmental growth represents an advance in terms of its
assessment of outcomes based on individual performance over time,
rather than cross-sectional analysis. The variable course of HIV
disease suggests that analysis of individual growth and change is
called for, with prompt identification of critical signs of
deceleration in developmental growth, rather than cross-sectional
analysis. Use of an extrapolation method for calculating MDI and PDI
scores <50 allowed inclusion of >98% of the sample of children with
3 or more assessments, thus more accurately reflecting the full spectrum of neurodevelopmental status. Standardized training of examiners, large sample size, use of a well-matched comparison group,
and multiple assessments within subjects over time are additional
important strengths of the study.
Ideally this study would have included a second control group of
HIV-unexposed children to control for the effects of maternal illness.
Other limitations include those imposed by the study design, including
the inability to control for use of antiretroviral treatment in
HIV-infected subjects and the low frequency of 2 SD drops in motor or
mental score as outcome events, which may have limited our ability to
detect small or subtle effects of other covariates. The assessment
instrument used was not designed as a measure of longitudinal growth
for individual children and may not adequately target specific aspects
of development that may be subtly impaired in HIV-1 infection.
Expanding the analysis beyond that of age-adjusted scores to include
analysis of individual children's raw scores on the BSID would have
allowed identification of children who lost previously acquired skills,
or who failed to gain any new skills over time, or who experienced an
insidious developmental decline that was not steep enough to be
identified in this study, but, nevertheless, may have represented
central nervous system disease progression. In addition, there
may be other covariates beyond those considered in our analysis,
including maternal smoking, severity of maternal illness, household
income, and developmental intervention services that contribute to
early neurodevelopmental growth or deficit in infants born to
HIV-infected women. However, taken together, these factors suggest that
the observed results may represent a minimum estimate of the size and
timing of the effect of HIV on early development.
The fact that exclusion of children with gestational age <34 weeks did
not substantially change the results is notable. Prematurity was
significantly associated with increased risk for MDI and PDI scores >2
SD below the mean (<69). However, prematurity seemed to have a
protective effect when confirmed drops of 1 SD were examined,
indicating the continued need to view carefully premature infants as a
subgroup in the selection and analysis of neurodevelopmental outcome
events. We chose to use prematurity as a covariate rather than using
developmental scores corrected for prematurity to evaluate the relative
contributions of HIV status and prematurity to neurodevelopmental performance. However, the use of developmental index scores that were
uncorrected for prematurity may have increased the probability that
premature infants would achieve an MDI or PDI <69 and may have
decreased the probability that premature infants would achieve a
confirmed drop in developmental index scores as defined in this study.
Although correcting for prematurity (subtracting the number of weeks to
term from the premature infant's postbirth chronologic age) is common
practice in scoring developmental tests for clinical use, for research
purposes, such practice hinders accurate assessment of the relative
effect of prematurity as an independent variable on dependent variables
of interest in multivariate analysis. Because the scaled scores of
premature infants were not corrected for prematurity (thereby
preventing over adjustment for this condition in multivariate analysis)
but scored according to the norms for actual chronologic age,
substantially lower MDI and PDI values were observed than if age had
been adjusted for prematurity.
Premature infants are less mature developmentally, and the use of
uncorrected scaled scores would be expected to increase the likelihood
of observing MDI or PDI scores <69. In addition, because the magnitude
of the effect of such adjustment (or lack of adjustment) is greatest
early in infancy, when the proportion of postbirth chronologic age that
the weeks to term correction factor represents is greatest, the use of
uncorrected scaled scores also would be expected to decrease the
probability that premature infants would experience confirmed drops in
MDI or PDI (by reducing early scores more than later scores). In the
case of 2 SD confirmed drops, however, no significant effect of
prematurity was observed, possibly attributable to the relatively
severe nature and smaller number of such events.
Current definitions of encephalopathy in pediatric HIV disease used for
epidemiologic surveillance include as one of the criteria for diagnosis
"the failure to attain or loss of developmental milestones or loss of
intellectual ability, verified by standard developmental scales or
neuropsychological tests."17 A clear working definition
of this criterion for clinical and research application has not yet
been formulated. Various indices of neurodevelopmental deficit and
decline might be selected to study the occurrence and timing of
abnormality in the developmental growth of HIV-infected infants and
children. These include plateaus or drops in raw scores, frequency of
abnormal age-adjusted scores, and drops in age-adjusted scores obtained
with standardized neurodevelopmental testing. To be useful, an index of
abnormality in neurodevelopmental growth related to HIV infection must
minimize the identification of neurodevelopmental deficit attributable
to variables other than HIV infection that are associated with risk for
deficits in cognitive and motor development, such as
prematurity,18 prenatal drug exposure,19,20
and poverty.21
The present study suggests that in the systematic diagnosis of HIV
encephalopathy in infants and children, a 32-point (2 SD) confirmed
drop in cognitive (MDI) or motor (PDI) scores on the BSID may be a
useful indicator of central nervous system progression. In
addition, the occurrence of MDI or PDI scores 2 or more SD below the
age-specific mean (<69) may be an even more useful indicator, but has
the potential to be confounded by effects of premature birth or
maternal educational attainment. Finally, 16-point (1 SD) confirmed
drops in MDI or PDI scores, although statistically more frequent in
HIV-infected subjects compared with HIV-uninfected subjects, lack
specificity for HIV disease and are additionally confounded by the
presence of other covariates, so that use of this particular outcome is
probably not appropriate in the application of current diagnostic
criteria for encephalopathy. The high frequency of 16-point confirmed
drops in scores among HIV-exposed but HIV-uninfected infants and
children must be viewed in the context of the socioeconomic stressors
that characterize the epidemiology of HIV infection in the United
States, the multiple demands placed on families caring for 1 or more
family members with life-threatening illness, and the frequent
separations of child and parent because of illness. In attempting to
explain the wide variability in individual developmental patterns in
infants born to women with HIV-1 infection, the possibility of early
encephalopathy, as evidenced by a deceleration of neurodevelopmental growth, must be weighed in the context of intervening environmental stresses and life events.
In contrast with findings of others who studied children exposed to
drugs but not to HIV,19-21 maternal use of illicit drugs
during pregnancy did not seem to play an important role in the genesis
of deficits and declines in cognitive or motor performance of infants
exposed to drugs and born to mothers with HIV infection.
Although cross-sectional analysis of mean scores at different ages may
have revealed differences between infants with and without prenatal
drug exposure, the relative severity of the neurodevelopmental outcome
events examined in this study and the observed powerful effects of HIV
infection and other covariates, such as premature birth, may have
diminished the ability to detect the effects of drug exposure.
| |
CONCLUSION |
Careful longitudinal monitoring of developmental growth is a
necessary component of comprehensive medical care for children with HIV
disease.22 One of the primary challenges in the care of
infants and children infected by HIV is the early and accurate
diagnosis of encephalopathy. BSID cognitive or motor scores 2 SD or
more below the mean (<69) likely are an important early marker of
severe HIV-related central nervous system disease. Likewise, 32-point
(2 SD) confirmed drops in MDI or PDI seem to be clear indicators of
early HIV-related central nervous system compromise. An even greater
challenge is the prompt and reliable identification of infants or
children at high risk for central nervous system disease progression,
before the onset of encephalopathy, to initiate early, aggressive
initiation of specific antiretroviral treatment in hopes of preventing
devastating effects of HIV on the developing brain. The association of
abnormal cognitive and motor developmental outcomes, defined by the
present investigation with declines in immune function, quantitative
measures of viral load, and progression of disease in other organ
systems, including the incidence and timing of deceleration of brain
growth and development of cortical atrophy, merits further study. Other quantitative and qualitative measures of developmental status and
changes also warrant investigation to assess their utility as early
indicators of risk for HIV-related central nervous system impairment in
infants with congenital or perinatal HIV infection.
| |
ACKNOWLEDGMENTS |
This study was supported by the National Institute of Allergy
and Infectious Diseases and the National Institute of Child Health and
Human Development, National Institutes of Health. It is also part of
the Women and Infants Transmission Study.
Principal investigators, study coordinators, program officers, and
funding include: Clemente Diaz, Edna Pacheco-Acosta (University of
Puerto Rico, San Juan, PR; Grant U01 AI34858); Ruth Tuomala, Ellen
Cooper, Donna Mesthene (Boston/Worcester site, Boston, MA; Grant UO1 AI
34856); Jane Pitt, Alice Higgins (Columbia Presbyterian Hospital, New
York, NY; Grant UO1 AI 34842); Sheldon Landesman, Hermann Mendez, Gail
Moroso (State University of New York, Brooklyn, NY; Grants HD-8-2913
and R0-1-HD-25714); Kenneth Rich, Delmyra Turpin (University of
Illinois at Chicago, Chicago, IL; Grant U01 AI 34841); William Shearer,
Celine Hanson, Norma Cooper (Baylor College of Medicine, Houston, TX;
Grant UO1 AI 34840); Mary Glenn Fowler (National Institute of Allergy
and Infectious Disease, Bethesda, MD); Vincent Smeriglio (National
Institute on Drug Abuse, Rockville, MD); and Sonja McKinlay, Leslie
Kalish, Susan Ellis (New England Research Institutes, Watertown, MA;
Grant N01 AI 35161).
We thank the investigators, study staffs, and especially the families
who have participated in the Women and Infants Transmission Study.
| |
FOOTNOTES |
Received for publication Dec 8, 1998; accepted Feb 22, 2000.
Reprint requests to (C.C.) Boston Medical Center, DLG 7, One
Boston Medical Center Pl, Boston, MA 02118-2393. E-mail:
cynthia.chase{at}bmc.org
| |
ABBREVIATIONS |
HIV-1, human immunodeficiency virus type 1;
AIDS, acquired immunodeficiency syndrome;
WITS, Women and Infants
Transmission Study;
MDI, Mental Developmental Index;
PDI, Psychomotor
Developmental Index;
BSID, Bayley Scales of Infant Development;
SD, standard deviation;
RR, risk ratio;
CI, confidence interval.
| |
REFERENCES |
-
Belman AL
Acquired immunodeficiency syndrome and the child's central nervous system.
Pediatr Clin North Am
1992;
39:692-714
-
Blanche S,
Tardieu M,
Duliege AM,
Longitudinal study of 94 symptomatic infants with perinatally acquired human immunodeficiency virus infection.
Am J Dis Child
1989;
143:1147-1153 [Abstract]
-
European Collaborative Study
Neurologic signs in young children with human immunodeficiency virus infection.
Pediatr Inf Dis J
1990;
9:402-406 [Medline]
-
Lobato MN,
Caldwell B,
Ng P,
Oxtoby MJ,
the Pediatric Spectrum of Disease Consortium
Encephalopathy in children with perinatally acquired human immunodeficiency virus infection.
J Pediatr
1995;
126:710-715 [CrossRef][Medline]
-
Centers for Disease Control and Prevention. HIV/AIDS Surveillance Report. Atlanta, GA: Centers for Disease Control and Prevention; 1994:1-391
-
Centers for Disease Control and Prevention. HIV/AIDS Surveillance Report. Atlanta, GA: Centers for Disease Control and Prevention; 1995:1-38
-
Wachtel RC,
Tepper VJ,
Houck DL,
Neurodevelopment in pediatric HIV-1 infection: a prospective study.
AIDS HIV Infect Fetus Adolesc
1993;
4:198-203
-
Nozyce M,
Hittelman J,
Muenz L,
Durako S,
Fischer ML,
Willoughby A
Effect of perinatally acquired human immunodeficiency virus infection on neurodevelopment in children during the first two years of life.
Pediatrics
1994;
94:883-891 [Medline]
-
Chase C,
Vibbert M,
Pelton SI,
Coulter D,
Cabral H
Early neurodevelopmental growth in children with vertically transmitted human immunodeficiency virus infection.
Arch Pediatr Adolesc Med
1995;
149:850-855 [Abstract]
-
Bayley N. Manual for the Bayley Scales of Infant Development. San Antonio, TX: Psychological Corporation; 1969
-
Naglieri JA
Extrapolated developmental indices for the Bayley Scales of Infant Development.
Am J Ment Defic
1981;
85:548-550 [Medline]
-
New England Research Institute. Women and Infants Transmission Study Procedures Manual, Version 4/19/91. Watertown, MA: New England Research Institute; 1991. Updates through January 29, 1993
-
National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research. The Belmont Report: Ethical Principles and Guidelines for the Protection of Human Subjects of Research. Washington, DC: Office of Protection from Research Risks; 1986
-
Code of Federal Regulations, Title 45 CFR Part 46. Protection of Human Subjects. Washington, DC: Office of Protections from Research Risks; 1992. Revised June 18, 1991
-
Centers for Disease Control and Prevention
1994 revised classification system for human immunodeficiency virus in children less than 13 years of age.
MMWR CDC Surveill Summ
1994;
43:1-15 [Medline]
-
Lindsey JC, McKinney RE, Propert KJ. Issues in the conduct of clinical trials for HIV-infected children. In: Finkelstein DM, Schoenfeld DA, eds. AIDS Clinical Trials. New York, NY: Wiley-Liss; 1995:267-286
-
Brady MT,
McGrath N,
Brouwers P,
Randomized study of the tolerance and efficacy of high- versus low-dose zidovudine in human immunodeficiency virus-infected children with mild to moderate symptoms (AIDS Clinical Trials Group 128).
J Infect Dis
1995;
173:1097-1106
-
Brooks-Gunn J,
Klebanov PK,
Liaw F,
Spiker D
Enhancing the development of low birth weight premature infants: changes in cognition and behavior over the first three years.
Child Dev
1993;
64:736-753 [CrossRef][Medline]
-
Griffith DR,
Azuma SD,
Chasnoff IJ
Three-year outcome of children exposed prenatally to drugs.
J Am Acad Child Adolesc Psychiatry
1993;
33:20-27
-
Neuspiel DR,
Hamel SC
Cocaine and infant behavior.
J Dev Behav Pediatr
1991;
12:55-64 [Medline]
-
Bendersky M,
Lewis M
Environmental risk, biologic risk, and developmental outcome.
Dev Psychol
1994;
30:484-494 [CrossRef]
-
El-Sadr W, Oleske JM, Agins BD, et al. Evaluation and Management of Early HIV Infection: Clincial Practice Guideline 7. Rockville, MD: Agency for Health Care Policy and Research, Public Health Service, US Department of Health and Human Services; 1994. Agency for Health Care Policy and Research publication 94-0572
Pediatrics (ISSN 0031 4005). Copyright ©2000 by the American Academy of Pediatrics
Top
Abstract
Methods
